DISEASES OF AQUATIC ORGANISMSDis Aquat Org

Vol. 97: 135–142, 2011doi: 10.3354/dao02395

Published December 6

INTRODUCTION

The intensive exploitation of salmonids is an activ-ity of high economic importance in Chile, this coun-try being second largest producer of Atlantic salmonSalmo salar in the world (Parada 2010). The produc-tion mainly takes place in the south of Chile and isdominated by the marine culture of Atlantic salmonwith an estimated production of ~400 000 t in 2007

(www.salmonchile.cl). However, after the outbreakof infectious salmon anaemia (ISA) in June 2007(Godoy et al. 2008), the Chilean aquaculture industryhad a reduction of 30% in total production during2009 compared to 2007.

In spite of the ISA outbreak, Piscirickettsia salmo-nis, the causative agent of piscirickettsiosis or sal -monid rickettsial septicaemia, is considered the prin-cipal agent affecting and causing mortalities in the

© Inter-Research 2011 · www.int-res.com*Corresponding author. E-mail: ravendano@unab.cl

Multiplex PCR for the detection of Piscirickettsia salmonis, Vibrio anguillarum,

Aeromonas salmonicida and Streptococcus phocaein Chilean marine farms

D. Tapia-Cammas1, A. Yañez2, G. Arancibia1, A. E. Toranzo3, R. Avendaño-Herrera1,*

1Laboratorio de Patología de Organismos Acuáticos y Biotecnología Acuícola, Facultad de Ciencias Biológicas, Universidad Andrés Bello, Viña del Mar, Chile

2Laboratorio de Enzimología, Instituto de Bioquímica, Facultad de Ciencias, Universidad Austral de Chile, Campus Isla Teja, Casilla 567, Valdivia, Chile

3Departamento de Microbiología y Parasitología, CIBUS, Facultad de Biología e Instituto de Acuicultura, Universidad de Santiago de Compostela 15782, Spain

ABSTRACT: A multiplex (m-)PCR-based protocol was designed for the simultaneous detection of themain marine bacterial pathogens in Chilean salmon farms: Streptococcus phocae, Aeromonassalmonicida, Vibrio anguillarum and Piscirickettsia salmonis. Each of the 4 oligonucleotide primerpairs exclusively amplified the target gene of the specific bacterial pathogen. The detection limit ofthe m-PCR using purified total bacterial DNA was 50 pg µl−1 for V. anguillarum, 500 fg µl−1 for P.salmonis, and 5 pg µl−1 for S. phocae and A. salmonicida. This corresponded to average limits in them-PCR sensitivity of 3.69 × 105 CFU ml−1 of V. anguillarum, 1.26 × 104 CFU ml−1 of S. phocae, and5.33 × 104 CFU ml−1 of A. salmonicida, while the detection limits for the spiked fish tissues, regardlessof the sample (spleen, kidney, liver or muscle) were 2.64 ± 0.54 × 107 CFU g−1 for V. anguillarum,9.03 ± 1.84 × 105 CFU g−1 for S. phocae, 3.8 ± 0.78 × 103 CFU mg−1 for A. salmonicida and 100 P. sal -monis cells. However, high amounts of DNA from 3 bacterial species had a reduction of ~1 log-unit onthe amplification sensitivity of S. phocae or A. salmonicida when these were present in lower concen-tration in the multiplex reaction. The assay described in this study is a rapid, sensitive and efficient toolto detect the presence of S. phocae, A. salmonicida, V. anguillarum and P. salmonis simul taneouslyfrom pure cultures and tissues from clinically diseased fish. Therefore, it may be a useful alternativeto culture-based methods for the diagnosis of infections in fish obtained from Chilean salmon farms.

KEY WORDS: Multiplex PCR · Fish pathogens · Atlantic salmon · Analytic sensitivity test · Specificity test

Resale or republication not permitted without written consent of the publisher

Dis Aquat Org 97: 135–142, 2011

marine Chilean farming every year with annual economic losses of US $100 million http://aqua.merck-animal-health.com/diseases/piscirickettsiosis/product additional_ 127_113333.aspx. Affected fish aredark in colouration, show inappetence, are lethargicand swim near the surface or in edges of the cages. In-ternally enlarged spleen, discoloured kidney andana emia, the principal characteristic of the disease,can be observed. In addition, some fish show distinctcircular nodules in the liver (see Fryer & Hedrick 2003for review).

In the last decade, incidence of other pathogenssuch as Streptococcus phocae (Romalde et al. 2008),Vibrio anguillarum (Silva-Rubio et al. 2008) andatypical Aeromonas salmonicida has increased inChilean salmon culture, provoking significant mor-talities also in rainbow trout Onchorhynchus mykissand Pacific salmon Oncorhynchus kisutch (Bravo2000, Godoy et al. 2010).

Each disease is diagnosed presumptively based onthe clinical signs of the affected fish. However, Streptococcus phocae, Aeromonas salmonicida, Vib-rio anguillarum and Piscirickettsia salmonis oftenappear in mixed infections, increasing the possibilityof misdiagnosis commonly attributed to P. salmonisperhaps due to the high incidence of the organism innet-cages. In order to prevent and control outbreaksassociated with mixed infections by these 4 patho -gens, development of an effective diagnostic methodis an important step.

The majority of Chilean laboratories employ isola-tion from fish tissues on agar media followed by iden-tification using biochemical and/or serological tech-niques. Until the development of an agar culturemedium (Mikalsen et al. 2007, Mauel et al. 2008), cell

culture in Chinook salmon embryo (CHSE) cells wasconsidered the gold standard to confirm Piscirick-ettsia salmonis (Lannan & Fryer 1991), but both cul-ture methods require intensive labour and are tech-nically demanding.

Polymerase chain reaction (PCR) has been used forrapid detection and identification of each of theabove microorganisms not only from plate culturesbut also in clinical samples from several fish organs(i.e. kidney, gill, heart, spleen and liver) without fur-ther isolation (Gustafson et al. 1992, Miyata et al.1995, Hirono et al. 1996, Mauel et al. 1996, 1999,Marshall et al. 1998, Gonzalez et al. 2003, Alber et al.2004, Nilsson et al. 2006, Avendaño-Herrera 2008,Beaz-Hidalgo et al. 2008, Hassan et al. 2008).

Nevertheless, it would be a relatively costly andlaborious process if single primer sets were used on alarge number of samples in individual PCRs (Altinoket al. 2008). An alternative to singleplex PCR wouldbe the use of simultaneous detection of fish patho-genic bacteria using a multiplex PCR (m-PCR)approach. In the present study, an m-PCR assay wasdeveloped for the simultaneous detection of Strep -tococcus phocae, Aeromonas salmonicida, Vibrioanguil larum and Piscirickettsia salmonis for the rapidand cost effective detection of these pathogens in fishfrom Chilean salmon aquaculture.

MATERIALS AND METHODS

Bacterial strains and growth conditions

The bacterial strains used to evaluate the m-PCRmethod in this study are presented in Table 1. This

136

Bacterial strain Source Isolation source No. of strains (AtS/RbT/CoS/ChS/Tb) detected

Isolated from diseased fishPiscirickettsia salmonis Laboratory collection 0/3/3/1/0 0/3/3/1/0Aeromonas salmonicida Laboratory collection 8*/0/0/0/3** 8/0/0/0/3Streptococcus phocae Laboratory collection 13/0/0/0/0 13/0/0/0/0Vibrio anguillarum Laboratory collection 1/0/2/0/3 1/0/2/0/3

Reference strainsStreptococcus phocae ATCC 51973T ATCC Phoca vitulina 1Aeromonas salmonicida subsp. salmonicida ATCC 33658 ATCC Salmo salar 1Aeromonas salmonicida subsp. achromogenes ATCC 33659 ATCC Salmo trutta 1Vibrio anguillarum ATCC 43307 ATCC Salmo gairdneri 1Piscirickettsia salmonis ATCC VR-1361 ATCC Oncorhynchus kisutch 1

Table 1. Strains used in the present study and results using the specific m-PCR detection method. AtS: Atlantic salmon; RbT:rainbow trout; CoS: Pacific salmon; ChS: Chinook salmon; Tb: turbot; ATCC: American Type Culture Collection (Rockville,

USA). *Atypical and **typical strains

Tapia-Cammas et al.: Multiplex PCR for salmonid pathogens

collection comprises 31 isolates from Atlantic salmonSalmo salar, Pacific salmon Oncorhynchus kisutch,rainbow trout O. mykiss and Chinook salmon O.tshawytscha in different Chilean salmon farms and 6isolates from turbot Psetta maxima. The type strainStreptococcus phocae ATCC 51973T, and the refer-ence strains Aeromonas salmonicida subsp. salmoni-cida ATCC 33658, A. salmonicida subsp. achromo-genes ATCC 33659, Vibrio anguillarum ATCC 43307and Piscirickettsia salmonis ATCC VR-1361 (equiva-lent to LF-89) were used as positive controls andwere obtained from the American Type Culture Collection.

In addition, reference strains of other pathogensfrom fish and mammals also obtained from outbreaks

were included to test specificity of the m-PCR(Table 2). The identity of each isolate was confirmedby standard phenotypical and microscopical tech-niques as reported by MacFaddin (1980) and also insome cases using PCR-based analysis or serologicalassays.

For all experiments, Aeromonas salmonicida andVibrio anguillarum strains were routinely grown ontryptone soya agar or broth supplemented with 1%(w/v) sodium chloride (TSA-1 or TSB-1, respectively)at 20°C for 24 to 72 h. Columbia sheep blood agarplates (CBA; AES Laboratory) were used to cultureStreptococcus phocae. For Piscirickettsia salmonis,BFCG agar (5% sheep blood agar plates that had theaddition of 3% fetal bovine serum [FBS], 0.1% cys-

137

Bacterial strain Source Detection by m-PCR

Vibrio spp. (5) isolated from diseased fish Laboratory collection –Vibrio ordalii (18) Laboratory collection –Vibrio ordalii ATCC 33509 ATCC –Vibrio harveyi TW425 Laboratory collection –Vibrio splendidus I CPV8.1 Laboratory collection –Vibrio splendidus II AZ206 Laboratory collection –Vibrio alginolyticus ATCC 17749T ATCC –Vibrio pelagius I TW487/02 Laboratory collection –Vibrio pelagius II RI 152.1 Laboratory collection –Aeromonas hydrophila 1404 Laboratory collection –Streptococcus parauberis RA9 Laboratory collection –Lactococcus garvieae TW94W Laboratory collection –Photobacterium damselae subsp. piscicida ATCC 29690 ATCC –Photobacterium damselae subsp. damselae AZ247.1 Laboratory collection –Pseudomonas anguilliseptica CECT 899 CECT –Tenacibaculum maritimum NCIMB 2158 NCIMB –Tenacibaculum maritimum NCIMB2153 NCIMB –Tenacibaculum ovolyticum NBRC 15947 NBRC –Tenacibaculum ovolyticum NBRC 15992 NBRC –Tenacibaculum mesophilum NBRC 16307T NBRC –Tenacibaculum mesophilum NBRC 16308 NBRC –Tenacibaculum amylolyticum NBRC 16310T NBRC –Tenacibaculum lutimaris DMS 16505 DMS –Rhodococcus qingshengii 79043–3 Laboratory collection –Flavobacterium psychrophilum 49418T ATCC –Flavobacterium columnare ATCC 23462T ATCC –Yersinia ruckeri CECT 955 CECT –Francisella sp. LM-84-F Laboratory collection –Chryseobacterium piscicola CECT 7357T CECT –Chryseobacterium chaponense Sa 1147-06T Laboratory collection –Hafnia alvei 15/1403 Laboratory collection –Enterobacter cloacae TW 03/03 Laboratory collection –Enterobacter aerogenes RPM799.1 Laboratory collection –Escherichia coli FV9180 Laboratory collection –

Table 2. Bacteria from other species included in this study as negative controls in the m-PCR analysis. NBRC: National Institute of Technology and Evaluation (NITE) Biological Resource Center (Osaka, Japan); ATCC: American Type CultureCollection (Rockville, MD, USA); NCIMB: National Collection of Industrial and Marine Bacteria (Aberdeen, UK); CECT:Colección Española de Cultivos Tipos (Universidad de Valencia, Spain). No specific amplification was detected for any strain.

Values in parentheses: number of strains assayed

Dis Aquat Org 97: 135–142, 2011

teine and 1% glucose) was used and incubated at16°C for 6 d (Mauel et al. 2008). Stock cultures werestored at −80°C in Cryo-bille tubes (AES Laboratory).

DNA extraction

Total DNA was extracted for subsequent analysisusing 2 different commercial systems: InstaGenePurification Matrix (Bio-Rad Laboratories) for pureand mixed bacterial cultures and AxyPrep™ Multi-source Genomic DNA Miniprep Kit (Axygen Bio-sciences) for tissue samples. In all cases, DNA purification was performed according to the manu-facturer’s instructions. The concentration and qualityof each DNA sample was examined spectrophoto-metrically at 260 nm on an EpochTM Microplate Spec-trophotometer. DNA from pure cultures was adjustedwith sterile distilled water to a concentration of 50 ±5 ng µl−1. Of each DNA sample extracted, 1 µl (10 to5 ng) was used directly for m-PCR amplification, andthe remaining DNA sample was stored at −20°C. Allexperiments were carried out with DNA obtained in2 independent extractions for each bacterial strainand isolate.

PCR amplification

All PCR amplifications were performed using aMastercycler personal (Eppendorf) and the commer-cial kit Ready-To-GoTM PCR beads (GE Healthcare)according to the manufacturer’s instructions. This kitincluded all the reagents needed for the PCR reac-tions (buffer, nucleotides and Taq DNA polymerase),with the exception of the specific primers and DNAtemplate. The sequences of the 4 PCR primer pairsfor m-PCR and the expected size of PCR productsgenerated with these primers are shown in Table 3.

The PCR annealing temperatures tested ranged from50 to 60°C. Intensity of the amplicons for each targetDNA, as well as the absence of nonspecific bands,was considered in selecting m-PCR conditions.Therefore, 55°C was used as the annealing tempera-ture for all PCRs. The cycling protocol was 1 cycle of95°C for 3 min, 35 cycles of 95°C for 1 min, 55°C for1 min and 72°C for 1 min, followed by a final elonga-tion at 72°C for 2 min.

Analysis of PCR products

Aliquots of 10 µl of PCR product were separated ona 1.5% (w/v) agarose gel for 60 min at 100 V in TAE1× (0.04 M Tris, 0.0001 M EDTA, pH 8.0) electro -phoresis buffer, visualized using 0.06 µg ml−1 of ethi dium bromide (Bio-Rad) and photographed un -der UV light. A 100 bp DNA ladder (100−1000 bp;Bioron) and GeneRulerTM 100 bp DNA Ladder Plus(100−3000 bp; Fermentas) were used as molecularmass markers. In all cases, negative controls consist-ing of the same reaction mixture, but with sterile dis-tilled water instead of template DNA, were includedin each batch of PCR reactions. The presence of a sin-gle product of the appropriate size, identical to therespective reference strains, was considered as apositive result.

Specificity and sensitivity from pure cultures

The specificity of the m-PCR was evaluated usingthe genomic DNA extracted from 37 related bacteriaand an additional 55 different fish pathogens or unre-lated bacteria from human (Table 2).

For the evaluation of the analytic sensitivity ofm-PCR, the minimum detectable amounts of tar-geted DNA by m-PCR were detected for Streptococ-

138

Bacterial species Target Primer Sequence (5’–3’) Amplicon Source gene size (bp)

Vibrio rpoN rpoN-ang5’ GTTCATAGCATCAATGAGGAG 519 Gonzalez et al. (2003)anguillarum rpoN-ang3’ GAGCAGACAATATGTTGGATG

Piscirickettsia 16S–23S RNA RTS1 TGATTTTATTGTTTAGTGAGAATGA 91 Marshall et al. (1998)salmonis RTS2 AAATAACCCTAAATTAATCAAGGA

Aeromonas vapA AP-1 GGCTGATCTCTTCATCCTCACCC 421 Gustafson et al. (1992)salmonicida AP-2 CAGAGTGAAATCTACCAGCGGTGC

Streptococcus 16S rRNA PX1 GCTAATACCGCATAAGAAGAG 900 Hassan et al. (2008)phocae PXVQ2 CACCCTGTCACTTCTGCTC Avendaño-Herrera (2008)

Table 3. Specific primers for m-PCR used in the present study

Tapia-Cammas et al.: Multiplex PCR for salmonid pathogens

cus phocae ATCC 51973T, Aeromonas salmonicidasubsp. salmonicida ATCC 33658 or A. salmonicidasubsp. achromogenes ATCC 33659, Vibrio anguil-larum ATCC 43307 and Piscirickettsia salmonisATCC VR-1361 strains as denoted by Onuk et al.(2010). After DNA extraction, serial dilutions of eachDNA were prepared with sterile distilled water atconcentrations ranging from 50 ng µl−1 to 500 fg µl−1.Aliquots of 10 µl of each dilution of DNA were mixedwith the respective aliquots of the other 3 strains.These DNA mixtures were used as template DNA forthe m-PCR experiments, and the detection limit forDNA was determined.

For the evaluation of analytic sensitivity of m-PCRfrom the cultures, based on bacterial cell numbers,separate bacterial suspensions of the Streptococcusphocae, Aeromonas salmonicida, Vibrio anguillarumand Piscirickettsia salmonis strains were prepared tocontain 109 cells ml−1 (McFarland Scale 4) and were10-fold diluted in 0.85% sterile saline solution from108 to 10 cells ml−1. Thus, bacterial mixtures with thedifferent fish pathogens were prepared and used asDNA template for m-PCR sensitivity testing. Chro-mosomal DNA was extracted as described previ-ously, and CFU ml−1 was estimated in all cases byplating onto agar plates. Limits of detection weredetermined based on presence or absence of PCRproducts on gels.

Determination of m-PCR sensitivity from spikedfish samples

M-PCR sensitivity was determined employingDNA extracted from in vitro spiked spleen, kidney,liver and muscle of juvenile healthy Atlantic salmonas described by Avendaño-Herrera et al. (2004).Each fish sample (mean weight: 16.3 to 12.2 mg) wasspiked with 100 µl of one of the bacterial dilutions(from 108 to 102 cells ml−1) from pure cultures of 4bacterial species and homogenized in phosphatebuffered saline (PBS, pH 7.4) for 60 s. After incuba-tion for 1 h at 18°C, DNA was extracted usingAxyPrep™ Multisource Genomic DNA Miniprep Kitaccording to the manufacturer’s instructions. Non-inoculated spiked samples, serving as negative con-trols, were processed in the same manner but withsterile saline solution instead of bacterial dilutions.For m-PCR, 1 µl of the purified DNA was added astemplate. Again, limits of detection were determinedbased on presence or absence of PCR products ingels.

The analytic sensitivity of the m-PCR was alsodetermined using spiked kidney tissue from Atlanticsalmon with different relative amounts of Streptococ-cus phocae, Aeromonas salmonicida subsp. salmoni-cida or subsp. achromogenes, Vibrio anguillarumand Piscirickettsia salmonis. Here, 100 µl of suspen-sion from each one of 3 bacterial species containing~107 CFU ml−1 was mixed with the same amount of asuspension of the fourth pathogen containing 104 to105 CFU ml−1. This low concentration was selectedconsidering the sensitivity obtained from the in vitrospiked tissues described above.

RESULTS AND DISCUSSION

Rapid diagnosis of outbreak agents is essential foreffective control, but current microbiological meth-ods based on culture and biochemical characteriza-tion are time-consuming. An m-PCR approach hasbeen successfully applied to detect multiple bacterialpathogens of marine (Kulkarni et al. 2009) and/orfreshwater fish (del Cerro et al. 2002, Mata et al.2004, Altinok et al. 2008, Altinok 2011).

Salmonid rearing in Chile has been seriouslythreatened by the appearance of a number of diversebacterial pathogens. In this investigation, an m-PCRmethod was developed to detect 4 marine pathogens,Streptococcus phocae, Aeromonas salmonicida, Vib-rio anguillarum and Piscirickettsia salmonis simulta-neously. Firstly, the analytic specificity of the 4primer sets was determined using the DNA extractedfrom 37 strains belonging to the 4 targeted species(Table 1), DNA from individual strains being addedto reactions. All 4 oligonucleotide primer pairs exclu-sively amplified the target gene of the specific bacte-rial pathogen, producing a unique and clear PCRproduct of the expected length for the correspondingmicroorganism (Fig. 1). With other strains or isolatesused in this study, neither specific nor false-positivebands were detected (Table 2).

The m-PCR condition was optimized for the simul-taneous detection of the 4 microorganisms, particu-larly the annealing temperature, and the best ampli-fication for all primer pairs being obtained at 55°C.The analytic specificity of the m-PCR assay was con-firmed using a DNA mixture prepared from the 4pathogens that gave amplification products of 900 bpfor Streptococcus phocae, 421 bp for Aeromonassalmonicida, 512 bp for Vibrio anguillarum and 91 bpfor Piscirickettsia salmonis (Fig. 1). No amplificationwas observed when DNA from other taxonomicallyand/or ecologically related bacteria were combined

139

Dis Aquat Org 97: 135–142, 2011

together, with the exception of V. splendidus andV. pelagius, giving a unique, non-specific (i.e. basedon size) PCR amplification product of weak intensity(unpubl. data). This can be explained by the high

degree of genetic relatedness of the Vibrio species,which is defined as a group of strains that share>95% DNA identity in multilocus sequence analysis(MLSA) (Thompson et al. 2009).

140

Fig. 1. Aeromonas salmonicida, Streptococcus phocae, Vibrio anguillarum, and Piscirickettsia salmonis. Amplification prod-ucts from m-PCR assay developed for the simultaneous detection of streptococcosis, typical and atypical furunculosis, vibriosisand piscirickettsiosis. MW: GeneRulerTM 100bp DNA Ladder Plus (100−3000 bp; Fermentas). Lanes 1 to 5: (1) negative control(no DNA), (2) S. phocae ATCC 51973T, (3) A. salmonicida subsp. achromogenes ATCC 33659, (4) V. anguillarum ATCC 43307,(5) P. salmonis ATCC VR-1361. Lanes 6 to 11: sensitivity of m-PCR in detecting the 4 fish pathogens from serially diluted (5 ngµl−1 to 50 fg µl−1) chromosomal DNA extracted from reference strains of each fish pathogen. Left y-axis: position of molecular

size marker (bp); right y-axis: size of specific amplified products (bp)

Fig. 2. Aeromonas salmonicida, Streptococcus phocae, Vibrio anguillarum, and Piscirickettsia salmonis. Detection by m-PCRof fish tissue samples containing a high amount of 3 pathogens (107 CFU ml−1 each bacterium) and low amounts of the fourthpathogen (104 to 105 CFU ml−1). CFU were estimated by standard plate counts. MW: Axygen 100bp DNA Ladder(100−3000 bp); Lane 1: negative control (no DNA); Lanes 2 to 5: DNA extracted from tissues spiked with (2) S. phocae ATCC51973T at 104 CFU ml−1, (3) V. anguillarum ATCC 43307 at 105 CFU ml−1, (4) A. salmonicida subsp. achromogenes ATCC 33659

at 104 CFU ml−1, and (5) P. salmonis ATCC VR-1361 (~150 fg µl−1)

Tapia-Cammas et al.: Multiplex PCR for salmonid pathogens

The analysis of sensitivity and robustness of the m-PCR protocol for the 4 species was determined byamplification of pure DNA of each reference strain induplicate. The range of detection using purified DNAwas 50 pg µl−1 for Vibrio anguillarum, 500 fg µl−1 forPiscirickettsia salmonis, 5 pg µl−1 for Streptococcusphocae and Aeromonas salmonicida, respectively(Fig. 1). In the case of DNA extracted from bacterialsuspensions, the average limits in the m-PCR sensi-tivity for each species were 1.26 × 104 CFU ml−1 of S.phocae, 3.69 × 105 CFU ml−1 of V. anguillarum and5.33 × 104 CFU ml−1 of A. salmonicida, while thedetection limits for the spiked fish tissues was thesame, regardless of the tissue type, with 9.03 ± 1.84 ×105 CFU g−1 for S. phocae, 2.64 ± 0.54 × 107 CFU g−1

for V. anguillarum, 3.8 ± 0.78 × 103 CFU g−1 for A.salmonicida and 100 P. salmonis cells.

Although the Piscirickettsia salmonis strain grewon BFCG agar (Mauel et al. 2008), poorly definedcolonies were produced, leading to inaccuracies inestimations of the CFU concentration. In fact, thebacterial count showed that the number of culturablebacteria decreased by 3 log-units (105 CFU ml−1) froman initial inoculum of 108 cells ml−1. This difficultyleads us to express the sensitivity of P. salmonis inDNA concentration units or cells only.

Taking into consideration our detection limit, thesimultaneous use of the 4 PCR primer pairs in onePCR resulted in a reduction in analytical sensitivityfor Aeromonas salmonida and Vibrio anguillarumcompared to that previously reported in the literature(Gustafson et al. 1992, Gonzalez et al. 2003). In thiscontext, our m-PCR procedure possesses similar de -tection limits for Streptococcus phocae and Piscirick-ettsia salmonis to those previous results obtained byAvendaño-Herrera (2008) and Marshall et al. (1998),respectively. In fact, Avendaño-Herrera (2008) hasreported for S. phocae detection level on the order of104 CFU ml−1 in bacterial suspension and 106 CFU g−1

in tissues, while for P. salmonis, Marshall et al. (1998)suggest that the simple PCR assay is capable ofrevealing the presence of 10 to 100 P. salmonis cells.The m-PCR procedure constitutes a powerful tool foraccurate identification of different pathogens fromplate cultures as well as from fish tissues (del Cerroet al. 2002, Mata et al. 2004, Altinok et al. 2008,Kulkarni et al. 2009, Onuk et al. 2010). However, animportant aspect is that this m-PCR is generallythought to be less sensitive than single PCR due tocompetition for reaction reagents among the assaysmaking up the multiplex technique, especially if theassays are different in their amplification efficiencyor one or more of the target organisms are present in

high numbers. In our investigation, the use of highamounts of DNA from 3 bacterial species (107 CFUml−1 of each pathogen) had a significant effect on theamplification sensitivity of the other bacterial speciesonly when A. salmonicida or S. phocae were presentin lower concentrations (104 to 105 CFU ml−1) in themultiplex reaction. In fact, the limits of detection ofthe S. phocae decreased by 1 log-unit (106 CFU g−1)when high amounts of amplification of the other 3target bacterial DNA occurred (Fig. 2). No amplifica-tion product was observed when DNA of A. salmoni-cida at lower DNA concentrations (105 CFU ml−1) wastested in the presence of higher DNA concentrations(107 CFU ml−1) of the other target species. We canspeculate that a possible cause can be due to theprimer length used and the nucleotide content. Ourresults are in agreement with Altinok (2011) whonoted that the high amounts of DNA from one bacte-rial species had a significant effect on the amplifica-tion sensitivity of the other bacterial species whenthese are present in lower concentration in the multi-plex reaction.

Nevertheless the level of sensitivity is probablysufficient to detect each pathogen in acute infectionsin fish. For the diagnosis of fish diseases caused byslow-growing bacteria such as Aeromonas salmoni-cida or Streptococcus phocae, PCR-based detectionis of particular importance because these bacteriacan be easily obscured by Piscirickettsia salmonisand are therefore likely to be underdiagnosed basedon cultivation only. Furthermore, m-PCR assay ismore rapid and cost-effective than singleplex PCR.In summary, this m-PCR can be an efficient tool todetect the presence of S. phocae, A. salmonicida,Vibrio anguillarum and P. salmonis simultaneouslyfrom pure cultures and tissues obtained from clini-cally diseased fish with high bacterial concentra-tions. It could therefore be a useful alternative to cul-ture-based methods for the diagnosis of infections inChilean salmon farms.

Acknowledgements. Funding for this study was provided inpart by Grant FONDECYT 1090054 from the ComisiónNacional de Investigación Científica y Tecnológica−CONI-CYT (Chile) and also by Grant DI-01-10/R from the Univer-sidad Andres Bello.

LITERATURE CITED

Alber J, El-Sayed A, Lämmler C, Hassan AA, Vossen A,Siebert U (2004) Determination of species-specific se -quences of superoxide dismutase A encoding gene sodAand chaperonin 60 encoding gene cpn60 for identifica-tion and phylogenetic analysis of Streptococcus phocae.Vet Microbiol 101: 117−122

141

Dis Aquat Org 97: 135–142, 2011142

Altinok I (2011) Multiplex PCR assay for detection of fourmajor bacterial pathogens causing rainbow trout disease.Dis Aquat Org 93: 199−206

Altinok I, Capkin E, Kayis S (2008) Development of multi-plex PCR assay for simultaneous detection of five bac -terial fish pathogens. Vet Microbiol 131: 332−338

Avendaño-Herrera R (2008) Simultaneous evaluation of fourPCR primer sets for the diagnosis of Streptococcus pho-cae infection. Dis Aquat Org 82: 217−222

Avendaño-Herrera R, Magariños B, Toranzo AE, Beaz R,Romalde JL (2004) Species-specific polymerase chainreaction primer sets for the diagnosis of Tenacibaculummaritimum infection. Dis Aquat Org 62: 75−83

Beaz-Hidalgo R, Magi GE, Balboa S, Barja JL, Romalde JL(2008) Development of a PCR protocol for the detectionof Aeromonas salmonicida in fish by amplification of thefstA (ferric siderophore receptor) gene. Vet Microbiol128: 386−394

Bravo S (2000) Occurrence of atypical furunculosis in Chile.Bull Eur Assoc Fish Pathol 20: 209−211

del Cerro A, Marquez I, Guijarro JA (2002) Simultaneousdetection of Aeromonas salmonicida, Flavobacteriumpsychrophilum, and Yersinia ruckeri, three major fishpathogens, by multiplex PCR. Appl Environ Microbiol68: 5177−5180

Fryer JL, Hedrick R (2003) Piscirickettsia salmonis: a Gram-negative intracellular bacterial pathogen of fish. J FishDis 26: 251−262

Godoy MG, Aedo A, Kibenge MJ, Groman DB and others(2008) First detection, isolation and molecular character-ization of infectious salmon anaemia virus associatedwith clinical disease in farmed Atlantic salmon (Salmosalar) in Chile. BMC Vet Res 4: 28

Godoy M, Gherardelli V, Heisinger A, Fernandez J and others (2010) First description of atypical furunculosisin freshwater Atlantic salmon, Salmo salar L., in Chile.J Fish Dis 33: 441−449

Gonzalez SF, Osorio CR, Santos Y (2003) Development of aPCR-based method for the detection of Listonella anguil-larum in fish tissues and blood samples. Dis Aquat Org55: 109−115

Gustafson CE, Thomas CJ, Trust TJ (1992) Detection ofAeromonas salmonicida from fish by using polymer -ase chain reaction amplification of the virulence sur-face array protein gene. Appl Environ Microbiol 58: 3816−3825

Hassan AA, Vossen A, Lämmler C, Siebert U, Fernández-Garayzabal JF (2008) PCR amplification of species-specific sequences of 16S rRNA and 16S-23S rDNA inter-genic spacer region for identification of Streptococcusphocae. Microbiol Res 163: 132−135

Hirono I, Masuda T, Auki T (1996) Cloning and detection ofthe hemolysin gene of Vibrio anguillarum. MicrobPathog 21: 173−182

Kulkarni A, Caipang CMA, Brinchmann MF, Kiron V (2009)Simultaneous detection of pathogens causing francisel-losis, furunculosis and vibriosis in Atlantic cod by multiplex

polymerase chain reaction. Aquacult Res 41: 1533−1538Lannan CN, Fryer JL (1991) Recommended methods for

inspection of fish for the salmonid rickettsia. Bull EurAssoc Fish Pathol 11: 135−136

MacFaddin (1980) Biochemical tests for identification ofmedical bacteria, 2nd edn. Waverly Press, Baltimore, MD

Marshall S, Heath S, Henriquez V, Orrego C (1998) Mini-mally invasive detection of Piscirickettsia salmonis incultivated salmonids via the PCR. Appl Environ Micro-biol 64: 3066−3069

Mata AI, Gibello A, Casamayor A, Blanco MM, DomínguezL, Fernández-Garayzábal JF (2004) Multiplex PCR assayfor detection of bacterial pathogens associated withwarm-water streptococcosis in fish. Appl Environ Micro-biol 70: 3183−3187

Mauel MJ, Giovannoni SJ, Fryer JL (1996) Development ofpolymerase chain reaction assays for detection, identifi-cation, and differentiation of Piscirickettsia salmonis. DisAquat Org 26: 189−195

Mauel MJ, Giovannoni SJ, Fryer JL (1999) Phylogeneticanalysis of Piscirickettsia salmonis by 16S, internal tran-scribed spacer (ITS) and 23S ribosomal DNA sequencing.Dis Aquat Org 35: 115−123

Mauel MJ, Ware C, Smith PA (2008) Culture of Piscirick-ettsia salmonis on enriched blood agar. J Vet DiagnInvest 20: 213−214

Mikalsen J, Olsen AB, Tengs T, Colquhoun DJ (2007) Francisella philomiragia subsp. noatunensis subsp. nov.isolated from farmed Atlantic cod (Gadus morhua L.). IntJ Syst Evol Microbiol 57: 1960−1965

Miyata M, Inglis V, Aoki T (1995) Rapid identification ofAeromonas salmonicida subspecies salmonicida by thepolymerase chain reaction. Aquaculture 141: 181−185

Nilsson WB, Gudkovs N, Strom MS (2006) Atypical strains ofAeromonas salmonida contain multiple copies of inser-tion element ISAsa4, useful as a genetic marker and atarget for PCR assay. Dis Aquat Org 70: 209−217

Onuk EE, Ciftci A, Findik A, Durmaz Y (2010) Developmentand evaluation of a multiplex PCR assay for simultaneousdetection of Flavobacterium psychrophilum, Yersiniaruckeri and Aeromonas salmonicida subsp. salmonicidain culture fisheries. J Vet Sci 11: 235−241

Parada G (2010) Tendencias de la acuicultura mundial y lasnecesidades de innovación de acuicultura chilena.Informe para el Consejo Nacional de Innovación para laCompetitividad, Santiago. http:// biblioteca. cnic. cl/ media/users/ 3/181868/files/18813/G_Parada_ACUI_final.pdf

Romalde JL, Ravelo C, Valdés I, Magariños B and others(2008) Streptococcus phocae, an emerging pathogen forsalmonid culture. Vet Microbiol 130: 198−207

Silva-Rubio A, Avendaño-Herrera R, Jaureguiberry B,Toranzo AE, Magariños B (2008) First description ofserotype O3 in Vibrio anguillarum strains isolated fromsalmonids in Chile. J Fish Dis 31: 235−239

Thompson CC, Vicente AC, Souza RC, Vasconcelos AT andothers (2009) Genomic taxonomy of Vibrios. BMC EvolBiol 9: 258

Editorial responsibility: Catherine Collins,Aberdeen, UK

Submitted: January 7, 2011; Accepted: July 8, 2011Proofs received from author(s): November 4, 2011